OBJECTIVES:  This study examines factors associated with health care costs among working age patients with diabetic neuropathic pain (DNP) who are undergoing duloxetine therapy.  METHODS:  Using a retrospective cohort design and claims data, we assessed the predictors of total healthcare costs over a one-year follow-up period for patients (18-64 years) with 30+ days of available duloxetine therapy between October 1, 2004 and December 31, 2005.  The index date was defined as the first observed pharmacy claim for duloxetine.  All individuals selected had 1+ diabetic neuropathy diagnosis and pain medication dispensed in the 12 months pre-index.  Two cohorts of patients were constructed for individuals with a starting duloxetine dose of 30 (“Dulox30”) and 60 (“Dulox60”) milligrams, respectively. Baseline demographic and clinical characteristics were assessed for each cohort, as was prior use of pain medications.  Resource utilization and expenditures were evaluated over 12-months following index.  Multivariate linear regression was performed to assess whether initial dose of duloxetine was correlated with healthcare costs, controlling for demographic and clinical characteristics.  RESULTS:  We identified 196 Dulox30 patients and 655 Dulox60 patients. The Dulox30 and Dulox60 groups were similar by age (55.0 vs 54.2, p=0.16), and gender (59.2% vs. 58.4% female, p=0.85).  A higher proportion of patients in Dulox30 group had 4+ diabetes-related complications (26.5% vs 18.4%, p<0.01) and a Charlson Comorbidity score >1 (23.0% vs 17.6%, p=0.03) relative to those in the Dulox60 group. Prior use of antidepressants was greater among Dulox60 patients whereas Dulox30 patients were more likely to receive oxycodone and hydrocodone (all p<0.01).  After adjusting for patient characteristics, patients in both cohorts had similar total healthcare costs (p=0.95). Factors associated with increased costs included advancing age, prior pain-related hospitalizations, insulin therapy, diabetes-related complications, and other chronic comorbidities.  CONCLUSIONS:  These findings indicate that one-year health care costs were similar for DNP patients receiving duloxetine therapy, regardless of initial dose.